Japan Review 25 (2013): 16984

REVIEW ARTICLE
Origins of Japanthe Big Picture Revisited:
A Review of New Plate Tectonics Research
Gina L. BARNES

This review essay mainly compares two articles by G.L. Barnes on Japanese
geology, previously published in Japan Review (2003, 2008), with a series of
articles on New Paradigms in Japanese plate tectonics published in Chigaku
zasshi in 20092010. The first purpose is to update and add new details to
flesh out the previous Japan Review overviews. A discussion about collisional
and accretionary tectonics then follows, outlining problems of interpretation
by scholars coming from different academic backgrounds (Alpine geology and
subduction-zone geology). This text is highly technical, based on the previous
offerings which should be read first.
Japanese geologists are forging ahead in determining new ways to
measure and interpret geological processes in a subduction zone. The Japanese
archipelago, composed of twenty seven geological belts, is affected by
movement of four different plates: two oceanic plates subducting under the
main islands, and the islands themselves apportioned between two continental
plates. The 500 million year history of the formation of the Japanese landmass
is of great general and theoretical interest but not well covered in formal textbooks. Thus, scientific papers such as the Chigaku zasshi offerings in Japanese
as well as those in English published in the prominent geology journals must
be synthesized to gain an understanding of this region. Since these subduction-zone movements have given rise to modern volcanoes and earthquakes,
that understanding forms a crucial background for disaster management.
New research mentioned herein includes zircon-dating of sediments in
accretionary complexes, identification of second continent formations in
the mantle, and tectonic erosion/accretion alternation.
Keywords: Japanese geology, plate tectonics, palaeogeography, structural
geology, accretionary complex, collision tectonics, tectonic erosion
Introduction
This author has previously published two articles in Japan Review that recounted the formation of the Japanese islands.1 Beginning around 2007, several articles have since appeared
1 Barnes 2003, 2008.

169

Gina L. BARNES

in scientific journals that are again changing our understanding of the geological origins
of Japan. In addition to reviewing this new research, my self-appointed task is to assess
whether my previous coverage of this topic remains accurate in light of these new findings.
In short, I can say yes, but several points need modifying. In addition, I offer a brief consideration of the overall problem of accretionary versus collision tectonics. Readers unfamiliar
with geological terminology are advised to read the earlier Japan Review articles, which
have extensive explanations of fundamental concepts plus glossaries for particular terms.
The present review builds on these earlier presentations, even as it amends the information
contained therein.
Here I survey a series of thirty five articles published in three parts under the title
Nihon rett keiseishi to jisedai paradaimu (Geotectonic
evolution of the Japanese Islands under New Paradigms of the Next Generation) in three
issues of the Chigaku zasshi 119:2 (2010), 119:6 (2010), and 120:1 (2011). All
articles but one are in Japanese with English titles and abstracts, and most illustrations
have captions in both Japanese and English. Free PDFs of these articles are available via the
Chigaku zasshi website, or directly from J-Stage (Japan Science and Technology Agency).
The English abstract for each article gives a good indication of content, while every issue
has its own preface in Japanese, in addition to an overview in English by Kasahara et al.
(2010), cited in the bibliography below. References herein to articles in the Chigaku zasshi
(CZ) series other than the English overview will be cited only by CZ volume, issue and page
numbers, keyed to the dates of publication above. Several of the findings of these original
reports have been synthesized by Isozaki et al. (2010), which is drawn on here together with
further sources in English; these are footnoted by author and date as usual, and appear in
the bibliography below.
The Chigaku zasshi collected papers contain an overwhelming wealth of new findings,
new ideas, and new concepts as well as historical perspectivesin the form of original
papers, review articles and historical reviews. Overall, the authors emphasize the importance
of Pacific-type (or Miyashiro-type) orogenythat is, mountain building through subduction rather than through collision, the latter getting more press through the formation of
the Alps and Himalayas. They emphasize that during the first two billion years of earths
existence, continents were formed exclusively through such subduction orogeny.2 This is still
the main mechanism for continental growth, but this understanding is now tempered by
the concepts of tectonic erosion and second continent formation, the latter of which forms
the substance of the new paradigm.
The concept of Pacific/Miyashiro-type orogeny has been challenged by French
geologists since the 1980s, with their experience in Alpine geology. A new article by Charvet
continues to dismiss this concept in favour of Alpine-type collision orogeny in an island arc,
insisting on collision as a major factor in the formation of the Japanese landmass.3 Charvet
objects to several aspects of Pacific/Miyashiro-type orogeny, but presents his viewpoint as
argument by assertion. Much detail, both old and new, is included but nothing that clearly
supports his assertions. I cannot rebut his work in detail here, but I include below instances
of collision in Japan that show that the proponents of Pacific/Miyashiro-type orogeny have
2 Maruyama et al., CZ 120:1, pp. 115223.
3 Charvet 2013.

170

Origins of Japanthe Big Picture Revisited

an open mind about collisionas they are open to any new developmentsbut that they
demand a firm grounding in the data and not unsubstantiated assertions.
New Paradigms and Challenges
The newest understanding
In 2010, Japanese geologists announced to the world that they had identified the driving
force that makes the Pacific Ring of Fire so tectonically active.4 The key words here are tectonically active. They do not refer to the volcanic eruptions and earthquakes common in
the Ring of Fire but to rifting: particularly, what is called second continent formation and
its role in causing continental rifting. This finding links interaction between the subducting
slab of oceanic floor and the overlying continental plate in a kind of boomerang effect.
Previous Japanese research concentrated on identifying the nature and contents
of Accretionary Complexes (ACs): those materials scraped off the ocean f loor during
subduction and accreted, together with trench deposits, to the continental plate edge in one
form of accretionary tectonicsa major mechanism of continental growth. In contrast, the
new research proposes that more often than not, the ocean floor grinds away the edges of
continental plates through tectonic erosion, rather than adding to them through accretion.
The eroded materials contain much granite previously produced through subduction zone
magmatism; these granitic materials are carried down by the subducting slab to between
520 and 660km deep, where they equilibriate and consolidate to form second continents.
The heat from naturally radioactive elements in granite in second continents causes the
formation of magma plumes in the overlying mantle, which rise to break up the earths
surface crust. These are much larger than the magma extrusions produced through any
individual volcano, and these plumes act especially to break up supercontinents. The many
rifted basins in the East Asian region, such as Lake Baikal and the South China, Philippine,
Bohai and Japan Sea basins, are thought to be caused by such superplume activity
from the second continent region.
Thus, subduction is not just something that happens at the edge of continents:
subducted materials affect the very coherence of the continents that overlie them. However,
whether the second continent revelation constitutes a new paradigm within plate tectonics
research is arguable, though the authors can certainly claim credit for making many aspects
of subduction and riftingand indeed their linkagemore understandable through this
model. In time we might see the second continent model join the long stream of Japanese
contributions to plate tectonics through identifications of the earthquake zone along the
downgoing oceanic slab (now called the Wadati-Benioff Zone) by Wadati in 1935, of paired
metamorphic belts by Miyashiro in 1961, of the Volcanic Front concept by Sugimura
in 1965, of Pacific-type orogeny by Matsuda and Ueda in 1971, and of the Accretionary
Complex concept by Kanmera in 1976.5

4 Kasahara et al. 2010.

5 As listed in Isozaki et al., CZ 119:6, pp. 9991053; see also Isozaki et al. 2010.

171

Gina L. BARNES

Revisions to understanding
Additional new information can be highlighted with reference to Isozaki et al., who have
identified several mistakes in past research that are best forgotten.6 These mistakes are
given in Table 1 (left hand column) along with the revised understandings (right hand
column). Many of those corrections will be revisited in the section below on Updates.
X Mistaken research results X

O Corrections O

XThat all of Japan formed and developed off the OInstead, most of the Japanese landmass formed off
North China craton
the South China craton
OInstead, the sub-horizontal faults separating the
XThat the most important faults are the Median
various Accretionary Complexes (ACs) are the major
Tectonic Line (MTL) and Fossa Magna
structuring agents of the Japanese landmass
XThat the origin of the MTL can be traced to a CreO The Pa leo- a nd Neo-tectonic lines operated
tacous strike-slip fault running along the continental
differently at different times, with only the latter being
edge and connected with the Tanakura Tectonic Line
a shallow strike-slip fault related to the TTL
(TTL)
OInstead, the Japan Sea basin rifted along two strikeXThat the Japan Sea opened (like French doors) from
slip faults, E and W, and the Japanese landmass moved
the middle, anchored at fixed points at the sides
directly away from the continent
XThat the most important faults in Japan are steep OInstead, the most important are the sub-horizontal
angle normal faults
faults separating ACs
XThat Japans several strike-slip faults are over
OInstead, the strike-slip faults are relatively shallow,
1000km long and some parts of Japan moved up along
recently activated, and affect the local landmass
them from the area of Vietnam
OInstead, it is composed of sub-horizontal thrust
XThat Japan is composed of strike-slip fault bounded
sheets that formed successively in similar subduction
terranes that do not share a common genesis
environments
XThat when ACs were not forming, the oceanic plate OThe absence of ACs is more likely due to loss of ACs
was subducting very obliquely
through tectonic erosion
X That the growth of ACs occurred continuously
OACs formed intermittently
seaward
OInstead, arc crust can be destroyed through tectonic
XThat once arc crust is formed, it is not destroyed and
erosion, and rather than growing in size, the landmass
continental growth continues seaward
can be reduced in size
XThat when an arc collides with an orogenic zone, OGranite basements of arcs can be subducted, as is
continental crust is not increased
currently occurring with the Izu Arc
Table 1. Revisions in Japanese plate tectonics research: mistakes as drawn from Isozaki et al. (CZ 119:6, pp.
9991053) with corrections given by the author. X (batsu) indicates wrong, O (maru) indicates correct.

Exportable research results

Isozaki, Maruyama and Yanai argue that Japanese geology is now in its latest stage of development: that of exporting science (after non-science, colonial science, and independent
science), explained as the ability to provide new indigenous research technologies to other
countries.7 One of these exports is using U-Pb (radioactive uranium decaying into stable
6 Isozaki et al., CZ 119:6, pp. 9991053.
7 Isozaki, Maruyama and Yanai, CZ 119:2, pp. 37891.

172

Origins of Japanthe Big Picture Revisited

lead) dating and LA-ICP-MS (Laser Ablation Inductively Coupled Plasma Mass Spectrometry) analysis of zircon crystals in Accretionary Complexes to reconstruct the subduction
record.8 Zircon crystals reside in the AC protolith (that is, AC sediments/sedimentary rocks
before being metamorphosed) and igneous rock fragments incorporated into the AC; some
zircons are detrital because they have come from already eroded and disappeared granite
bodies and sandstones that have been recycled into these AC. By investigating where they
came from, the nature of the inner continent behind the subduction zone where the AC was
forming can be ascertained.
A second export involves using seismic profiling to reconstruct both second continent
regimes and the three-dimensional nature of ACs.9 Germane to this is the need to establish
the importance of faults according to their structural roles, with tectonic unit boundaries
taking priority over other faults which are perhaps more recent, more visible, and account
for current re-arrangements rather than fundamental geotectonic arrangements. In assessing
the relationship between geological belts, it is necessary to recognize the three-dimensional
forms of regional metamorphic zones. This allows assessment of continuity in relationships
between adjacent belts from unmetamorphosed ACs through to High Pressure/Low Temperature metamorphic belts.
Finally, the revised view on granite destinations is exportable (Isozaki et al. 2010):
namely that arc batholiths can disappear through tectonic erosion, leading to second
continent formation; that tectonic erosion contributes to the formation of serpentinite
belts (transformed mantle rock), which are later exhumed to the surface; and that there is
significant subduction of buoyant continental crust in the Izu-Bonin Arc.
Updates to The Big Picture
In the following section, the new Japanese research will be compared with what was published
in Japan Review in Barnes 2003 and 2008. The dating conventions are: Ga = billion years
ago, Ma = million years ago.
I reported that accretionary tectonics have been a relatively unknown compositional
aspect of the Japanese islands by non-geologists, who usually think of Japan as a volcanic
archipelago. In fact, the geological belts of the Japanese landmass are almost entirely
composed of ACs with some continental fragments, but many overlie and are intruded by
igneous masses. Worldwide, the survival of Accretionary Complexes is rare, making Japan
an exemplary case. The more common process is the loss of continental and oceanic crust
through subduction-erosion.10 ACs are estimated to have formed during one third of the
time since subduction began, while erosion occurs during the other two thirds;11 today, 75%
of subduction zones around the world are eroding, while 25% are accreting.12 The Japanese
islands therefore are valuable resources for studying ACs as phenomena of crustal growth,
especially in a world where continental crust is now decreasing in mass.13
8 Isozaki et al., CZ 119:6, pp. 9991053; Nakama et al., CZ 119:6, pp. 116172.
9 Isozaki et al. 2010.
10 Maruyama et al., CZ 120:1, pp. 115223; Tsujimori, CZ 119:2, pp. 294312; Ueda, CZ 119:2, pp. 36277.
11 Suzuki et al., CZ 119:6, pp. 117396.
12 Yamamoto, CZ 119:6, pp. 96398.
13 Yamamoto, CZ 119:6, pp. 96398.

173

Gina L. BARNES

ACs which do not survive are tectonically eroded from the over-riding plate by the subducting platealong with granitic continental crust. These materials are taken down into
the mantle where they are understood to form second continents.14 It is estimated that five
to seven times the amount of continental crust that exists on the earths surface is contained
in these submerged second continents.15 Isozaki et al. propose that accretion and erosion
along the continental edge occurred alternately.16
I took pains to emphasize that the present day Japanese Islands have only existed as islands
since ca. 15 million years ago. It now appears that at least five series of Japan arcs (defined
by granite basements) have been created, among which some proto-Japan arcs have existed
as continental arcs and others as archipelagos similar to modern Japan.17 Three previous arcs
have been subducted and lost to second continent formation, except for blocks of granite
and metamorphic rocks incorporated intoand therefore surviving inexisting serpentinite mlanges.18 Serpentinite is the mantle rock peridotite after it has absorbed water, thus
becoming soft and acting as a trap for other rocks pushed into it (forming a mlange) as the
downgoing slab scrapes against the mantle.
The initiation of subduction in the Japan area has been pushed back from 450Ma to
520Ma.19 U-Pb dating of zircons has revealed clusters of dates when granite batholiths
underlying arc formations in subduction zones were created at 520400Ma, 280210Ma,
and 190160Ma; these have all been eroded away except for current small exposures.20 The
mi serpentinite mlange in the Hida Marginal Belt (Hida Gaien) represents the
earliest subduction phase,21 while several other serpentinite belts contain granite fragments
dating to 450Ma and 250Ma. 22 The Cretaceous batholiths, dating to 11090Ma and
8060Ma still exist in the Sany and Sanin districts, respectively.23
Further zircon datings reveal seven periods of continental erosion whose materials were
eventually recycled into the proto-Japan area: 25001000Ma (Paleo- to Mesoproterozoic),
1000800Ma (Neoproterozoic), 520400Ma (Cambrian-Silurian), 280210Ma (Permian-Triassic), || 190160Ma (Jurassic), 11090Ma (mid-Cretaceous), || and 8060Ma (Late
Cretaceous-Paleogene).24 The double lines here indicate when there was a major change
in the source of the erosional materials: the first at 200Ma when the North and South
China cratons collided, and then at ca. 80Ma after the formation of the Cretaceous granite
batholith, which kept earlier continental materials from flowing into the proto-Japan area.

14
15
16
17
18
19
20
21
22
23
24

174

Sensh, Maruyama and Rino, CZ 119:6, pp. 121527.

Kawai et al., CZ 119:6, pp. 11971214.
Isozaki et al. 2010, pp. 95100.
Isozaki et al. 2010, p. 102, Figure 12-c.
Isozaki et al., CZ 119:6, pp. 9991053.
Kunugiza and Got, CZ 119:2, pp. 27993; Santosh and Sensh, CZ 120:1, pp. 100114.
Nakama et al., CZ 119:6, pp. 116172.
Matsumoto et al., CZ 120:1, pp. 429.
Suzuki et al., CZ 119:6, pp. 117396, Figure 5.
See Isozaki et al. 2010: Figure 6 cont.
Isozaki et al. 2010, p. 89.

Origins of Japanthe Big Picture Revisited

In Figure 5 and text of Barnes 2003, I represented the then current understanding that the
Hida Belt was affiliated with the North China craton (ancient continent) and the Oki
Belt derived from the South China craton. The dates of detrital zircons in the Oki Belt are
now understood to be as old as the Hida Belt, matching the date of the North China craton
at 2.01.8Ga when the craton resided in the center of the former supercontinent Nuna/
Columbia. Conversely, detrital zircons recovered in Japan from the South China block date
only to ca. 1Ga when that craton resided in the center of the next supercontinent, Rodinia.25
Thus both the Hida and Oki Belts are now thought to derive from the North China craton.
Upon the break-up of Rodinia, subduction began at ca. 520Ma, as represented in rocks
from the Hida Marginal Belt representing an early island arc. The oldest sedimentary unit, a
felsic tuff bed, has been identified at 472Ma, indicating the existence of explosive volcanoes
at that time.26 The next oldest sedimentary unit is the South Kitakami Belt, comprised of shelf sediments post-dating 444Ma.27 Other oldest units in the Hida Marginal
Belt are these: oldest ophiolite 520Ma, oldest granite 520Ma, oldest metasomatism 520Ma,
oldest blueschist 450Ma.28
One of the more fascinating aspects in studying the early geology of Japan is the reconstruction of paleogeographic maps. Those presented in Barnes 2003/2008 have now been
updated in Isozaki et al., with commentary (Section 7).29 These give a more dynamic view of
the units being discussed here.
The collision and unification of the North and South cratons is represented on the
continent by the Qinling-Dabie-Sulu suture, which covers a large extent in
China and crosses into the Korean Peninsula. Four tiny parcels of mid-temperature/pressure
metamorphic belts claimed to correlate with this suture zone have recently been found at
widely separated locations in Kyushu , near the Noto peninsula and in eastern
Thoku .30
I noted that the exhumation of high pressure/low temperature metamorphic belts was
a common phenomenon in subduction tectonics. There now exists a clear model for how
AC materials, subducted to depths that result in blueschist facies metamorphism, are then
squeezed up by insertion during non-accretionary, erosive periods.31 Moreover, the current
crystalline contents of these metamorphosed units represent not the peak pressure/temperature of metamorphism but a retrogressive crystallization as the unit dehydrated during
exhumation.32

25
26
27
28
29
30
31
32

Santosh and Sensh, CZ 120:1, pp. 100114; Nakama et al., CZ 119:6, pp. 116172.
Nakama et al., CZ 119:2, pp. 27078.
Shimoj et al., CZ 119:2, pp. 25769.
Isozaki et al. 2010, Figure 12.
Isozaki et al. 2010, Figures 1114.
mori and Isozaki, CZ 120:1, pp. 14051.
Ueda, CZ 119:2, Figure 9-B.
Isozaki et al. 2010, p. 89.

175

Gina L. BARNES

I referred to the Japanese archipelago as a mature island arc in contrast to the Izu Arc.
Six stages of continental crust growth are now recognized in Pacific-type orogeny, with
different parts of Japan used to represent stages 3, 5 and 6.33
Ophiolites, which are remnants of oceanic oor, are scattered throughout the Japanese
islands. These are likely to have been formed in back-arc or fore-arc positions around previous proto-Japans (similar to the existing oceanic floor in the Japan Sea formed 2015Ma)
rather than representing oceanic floor from the Pacific Ocean plate predecessors. The oldest
ophiolite in Japan, however, at 600Ma, is thought to derive from a passive margin setting of
oceanic plate (similar to the current Atlantic Ocean shores) before subduction began.34
The Sanbagawa Metamorphic Belt is shown as cutting through the middle of Kyushu
in Figures 7 and 9 of Barnes 2003. This belt has been discovered to incorporate at least two
different ACs, formed and metamorphosed at different times; the geographical extents of
these separate components have now been revised.35 The two parts are now referred to as
the Sanbagawa Metamorphic Belt sensu stricto and the Shimanto Metamorphic Belt
(MB). Rocks of the Shimanto MB have been discovered in northern Kyushu on the western
side of the Nomo Tectonic Line in Nagasaki prefecture, indicating substantial
displacement of south western Japan southwards along the fault, perhaps as part of the
Japan Sea opening.36
The Median Tectonic Line, one of the major constructional faults running through
western Japan, was earlier believed to have shifted Outer Zone strata (on the southern rim
of Japan) more than 1000 miles northwards from off the southern China coast to their
present position in Japan. This interpretation of the once popular but unrealistic strike
slip fault-controlled tectonics in SW Japan has now been discarded in favour of in situ
sub-horizontal stacking of ACs.37 The MTL is now discussed in two chronological senses:
the paleo-MTL, which was a mechanism of arc shortening during the opening of the Japan
Sea basin; and the neo-MTL, which is a Quaternary strike-slip fault that demarcates the
boundary of the Nankai fore-arc sliver.38
Because NE and SW Japan are subducting differently aged and differentially hydrated
plates (Pacific and Philippine, respectively), their volcanic eruption patterns and chemistries
are completely different.39
The history of the Izu Arc was previously presented as a collision with Honshu after
15Ma in five separate (intermittent) accretion events.40 The juncture of the Izu Arc with
Honshu, involving the Philippine Plate, the North American Plate (northeast Honshu) and
33
34
35
36
37
38
39
40

176

Maruyama et al., CZ 120:1, pp. 115223.

Isozaki et al., CZ 119:6, pp. 9991053.
Aoki et al., CZ 119:2, pp. 31332; Otoh et al., CZ 119:2, pp. 33346.
Kchi et al., CZ 120:1, pp. 3039.
Isozaki et al. 2010; It and Sat, CZ 119:2, pp. 23544.
Isozaki et al. 2010, p. 95.
Katayama et al., CZ 119:2, pp. 205223.
Hirata et al., CZ 119:6, pp. 112560.

Origins of Japanthe Big Picture Revisited

the Eurasian Plate (southwest Honshu), is now recognized as a triple trench junction.41 This
onshore triple junction is twinned with the offshore triple trench junction of the Philippine,
North American and Pacific Plates.
Recent seismic studies on the Izu Arc indicate that buoyant material such as an island
arc can be easily subducted if less than 25km thick.42 The bending of the geological belts in
the Kant syntaxis at 5Ma was ostensibly caused by crust greater than this thickness,
accompanied by the accretion of crustal blocks now marooned as the Tanzawa and
Misaka mountains. The majority of the Izu Arc, subducted since 17Ma, amounts to
about 700800km in length.
One of the reasons Izu Arc materials were intermittently accreted is the angle of the arc
meeting Honshu: perpendicular (orthogonal); when an arc collides obliquely, it subducts
smoothly, as is currently the case with the Kyushu-Palau Ridge subducting under Kyushu.43
In contrast, an intra-oceanic arc that did not subduct but accreted in the Triassic exists in
the Maizuru Belt.44
I reported that the Japan Sea opening took place between 19 and 15Ma; completion of
rifting is now dated to 16Ma.45 The whole rifting process has now been extended to include
a series of processes: a superplume causing arc volcanism at 3937Ma and resulting in
general uplift; the beginning of normal faulting and seawater intrusion ca. 33Ma; rifting
far enough to begin ocean floor formation ca. 20Ma; and further arc volcanism between
2015Ma accompanying rifting.46
I also reproduced Matsuda and Otofujis 1984 palaeomagnetic data in understanding
the opening of the Japan Sea as NE and SW Japan rotating down from the continent
like French doors opening outwards.47 This model is now rejected in favour of movement
(like an opening drawer) directly away from the continent along two transform faults: the
Tanakura Tectonic Line in Thoku and the Ululun Tectonic Line running down the
west of Kyushu;48 this movement was accomplished with very little rotation, leaving SW and
NE Japan microplates as the largest of thirty one separate continental blocks rifted from the
continental edge.49 Yanai, Aoki and Akahori (Figure 7) propose a reconstruction that packs
these blocks back into the area of the current Japan Sea, where many remain marooned
there today. The portions of north eastern and south western Japan were widely separated,
while the Outer Zone is shown in approximate position circa 2025Ma. They estimate
that 150km of crust separated the current Inner and Outer Zones and was lost (via thrust
movement) when these were brought into their present positions.

41
42
43
44
45
46
47
48
49

See Isozaki et al., CZ 199:6, Figure 2.

Yamamoto et al. 2009.
Isozaki et al. 2010, p. 95.
Isozaki et al. 2010, Figure 13-c.
Nakama et al., CZ 119:6, pp. 116172.
Yanai, Aoki and Akahori, CZ 119:6, pp. 10791124.
Barnes 2003, Figure 8b.
Isozaki et al. 2010, p. 95.
Yanai, Aoki and Akahori, CZ 119:6, Figure 3.

177

Gina L. BARNES

Geotectonic Belt Descriptions

The tectonic units that make up the Japanese landmass are traditionally called belts (tai
); elsewhere these may be termed terranes, but the imported term terrane is avoided
because it too often implies an exotic (allochthonous) origin and is testimony to the nature
of colonial geology in Japan.50
Figure 1 presents a compilation of information on these various tectonic belts, several
newly identified; equivalencies are clarified for those tectonically rearranged belts which
have different names in different parts of the country but are of the same origin. Other more
minor belts appear in the Japanese geological literature but are not included here. Notes to
the map include the caution that the granite belts and their contact-metamorphic aureoles
(Ryke/Gosaisho /) are very different from most other belts, which began as
Accretionary Complexes and have undergone varying degrees of regional metamorphism.
The continental fragments (Hida, Oki Belts) are also of a completely different origin than
the ACs that make up the majority of the Japanese landmass. Finally, the South Kitakami
Belt is noted for being the only example of very thick ancient shelf sediments.
All of these geotectonic belts are fault-bounded. Considerable work has been expended in identifying the kinds of faults that separate each type of unit and how important
they have been in constraining the developmental sequence of the Japanese landmass.51
Interestingly, the big, recently active faults such as the Median Tectonic Line (neo-MTL),
Itoigawa-Shizuoka TL, Fossa Magna, Kant (Tone ) TL, and the
Tanakura TL are classified as steep-angle normal strike-slip faults of minor fourth class
significance, as they are boundaries of microplates which were activated relatively recently in
the opening of the Japan Sea.
The more significant class one and two major faults are low-angle normal or reverse
(thrust) faults that occur between the continental craton/suture fragments and an AC, and
between an AC and a high-pressure metamorphic belt. Because the faults are low-angle, the
relationship between neighbouring belts is sub-horizontal, meaning that faults define the
tops and bottoms of stacked units. The geological belts are thus layers, not blocks (as they
appear to be in surface outcrop on a map).
A full N-S cross-section of the archipelago has been constructed through Shikoku ,
Chgoku and the Oki Ridge, showing the sub-horizontal layering in the Inner
Zone, and a slightly oblique layering in the Outer Zone.52 The MTL has traditionally been
considered the boundary between these two zones, but this needs clarifying following the
separate identifications of the paleo-MTL and neo-MTL: it is the paleo-MTL that separates
the Ryke granites of the Outer Zone from the Sanbagawa Metamorphic Belt of the Inner
Zone.53 The major Tectonic Lines that structure the Japanese landmassfrom oldest to
youngest, north to south: Nagato-Hida, sayama , mi , Ishigaki-Kuga, Butsuz, and Aki are essentially sub-horizontal thrust faults which represent previous
subduction planes.54

50
51
52
53
54

178

Isozaki et al., CZ 119:6, pp. 9991053.

Isozaki et al., CZ 119:6, pp. 9991053; Yanai, Aoki and Akahori, CZ 119:6, pp. 10791124.
It and Sat, CZ 119:2, pp. 23544.
Isozaki et al., CZ 119:6, pp. 9991053.
Isozaki et al. 2010, Figure 8.

Origins of Japanthe Big Picture Revisited

Figure 1. Descriptions of Geological Belts comprising the Japanese landmass (condensed from Isozaki et al., CZ 119:6,
pp. 10131022; Isozaki et al. 2010, pp. 10131022). Tectonic Line (TL) abbreviations: BTL = Butsuz , I-KTL
= Ishigaki-Kuga , MTL = Median, Ng-HmTL = Nagato-Hida Marginal . Other abbreviations:
TTT = triple trench junction; IP = igneous province; mJ-eK = middle Jurassicearly Cretaceous; lK-Pg = late CretaceousPalaeogene.

Collision Orogeny in Japan?

Charvet states that the first and main difficulty with the Pacific-Miyashiro Orogeny model
is that it is based on ridge subduction causing nappe (AC) emplacement; he argues that ridge
subduction causes tectonic erosion, not accretion.55 Instead, he insists that the arrival at the
trench of a small block with a lighter crust, and therefore positive buoyancy, like a mature arc
or micro-continent is able to induce a collision producing permanent compressive structures.56
It seems to me that Charvet has misunderstood the causal sequence of the proposed
role of ridge subduction in accretion. Figure 3 in Isozaki et al. (2010) clearly shows that
ridge subduction is followed several million years later by AC formation. Isozaki and
colleagues do not deny the reality of tectonic erosion at the continental edge, but as I
understand it, they argue that the subduction of a hot oceanic ridge will cause both uplift
and magma production at the continental edge. The uplift and increased volcanic activity
will result in a higher rate of erosion, which supplies sediments to the trench. It is these
sediments that are then bulldozed back up into the continental edge to form an AC, long
after the ridge has been subducted.
55 Charvet 2013, p. 92.
56 Charvet 2013, p. 94.

179

Gina L. BARNES

Belt

Abbreviation

Nature

Dates
(those from
detrital zircon
in parentheses)

Protoliths of metamorphic belts; Contents of

others

Metamorphic
Metamorph.
Notable rock types
facies P = pressure
Date
(minor)
T = temperature

Pre-Cambrian
(3.4Ga, 2.6Ga,
gneiss; Jurassic
Hd
280-240Ma
2.0-1.7 Ga,
granite intrusions

1.1Ga, 580Ma)
Oki
continental
Pre-Cambrian
gneiss; Jurassic
Ok
silt, sand
280-240Ma medium
fragment
(3Ga, 2Ga, 1.7Ga)
granite intrusions

serpentinite
metamorphosed
lowT/highP
400Ma granitoid
Nagatomlange,
component from
crystalline schists
gneiss as tectonic
Renge
NR=Kr includes Hida Palaeozoic
Ordovician~Devonian 450-340Ma occur as tectonic
blocks in western
Gaien &
AC; includes ultrabasic
blocks in eastern

part
Renge schist
rock
part
MiyamoriOphiolite +
Ordovician
highP rocks in
Hayachine MH=NR serpentinite
(480-470Ma)
mlange
mlange
ophiolite

old name,
Sangun
now split into
Sn
Renge and Suo

schists
includes
serpentinite
early~late
same origin as NR
mlange
Palaeozoic
Kurosegawa
and pre-Jurassic
Kr=NR existing as a
volcanics,
serpentinite mlange

klippe in
sediments &
beneath it
Chichibu Belt
metamorphic
rocks
continental
fragment

Hida

S. Kitakami

Nedamo

Higo
+ locs. In
Unazuki
,
Hitachi
HitachiTakanuki

Akiyoshi

Suo

Maizuru

180

Sk

Nd

Hg

thick shelf
sediments
Accretionary
Complex;
Permian part
same as Ak

post-OrdovicianMesozoic

QinlingDabieshanSulu suture
fragments

(3.71.9Ga
detrital zircons)
in 260230Ma
granites

northeastern
HT (Htfragment of
Tk) = Hg
Hg

440410Ma arc
granites

CarboniferousPermian

silt, sand, fossiliferous

limestone, Cambrian 230Ma
volcanics

medium

Cambrian and
CarbonimedP
ferrous sediments,
volcanics
sandstone, mudstone
chert, reef limestone,
Late Permian weak
seamount greenstone
and ocean floor basalts
lowT/highP
silt, sand, Permiangreenschist240Ma
Triassic AC
glaucophane
schist facies

Ak

Accretionary
Complex

Su

Metamorphic
Belt

Mz

fragment
Permian granite,
of Palaeoophiolite;
zoic island arc; Permian-Triassic
Permian-Triassic
closely related
sandstone
to UT

PermianCretaceous
granites

Origins of Japanthe Big Picture Revisited

Belt

UltraTanba

Chizu

Abbreviation

Sanbagawa

Cz

separated from
Metamorphic
Suo Belt by
Belt
zircon dates
Accretionary
Complex
Accretionary
Complex
Accretionary
Complex
Accretionary
Complex

granite
batholith and
Ry = Gs
metamorphic
aureole
Sb
sensu
stricto

Southern
Shimanto

Shm

Sh-S

Northern
Shimanto

Hidaka

Nemuro

Metamorphic
Metamorph.
Notable rock types
facies P = pressure
Date
(minor)
T = temperature

Permian-Triassic AC

weak

Triassic AC

lowT/highP
greenschistglaucophane
schist facies

180Ma

silty sandy
greenschist

end of TriassicJurassic

early Cretaceous seamounts

Jurassic

Cretaceous

M-T Jurassic AC +
Ak Permian AC +
Su Triassic highP
metam. + Hg medP
metamorphics

<80Ma protolith
(part of Sh-N that mid-Cretaceous AC
metamorphosed)

Cretaceous

lower mylonite

mid-Creta. lowT/highP
120110Ma greenschistpeak
eclogite facies

silty sandy
greenschist,
covered w/
eclogite schist

lowT/highP
term. Creta.
greenschist60Ma peak
eclogite facies

silty sandy
greenschist

Cretaceous

lowT/highP

Cretaceous

lowT/highP

Jurassic AC

weak

Palaeogene AC,
southern end is
Miocene AC

unmetamorphosed AC

unmetamorphosed AC
mlange
Accretionary
Hdk = Sh
Complex
shelf
Nm
sediments
Sh-N

Protoliths of metamorphic belts; Contents of

others

Metamorphic Cretaceous,
early Cretaceous AC
Belt
140Ma protolith

Metamorphic
Belt divided
from Sb s.s.
Kamui kotan
Metamorphic
Kk = Sb
Belt

Tokoro
Metamorphic
Tk = Sb
Belt

granite
Gosaisho
batholiths and
Gs = Ry
metamorphic

aureole
Shimanto

Dates
(those from
detrital zircon
in parentheses)

UT

MinoTanba M-T =
Ch

Chichibu outlier of
M-T

Sanpzan subset of
Chichibu

N. KitakamiNk-Os =
Oshima
M-T, Ch

Ryke

Nature

<80Ma

late Cretaceous

Cretaceous

Cretaceous AC

sandstone
mudstone
(chert greenstone)
sandstone
mudstone
(chert greenstone)
Cretaceous

partially metamorphosed

Cretaceous

Table 2. Geological belts of Japan (modified from Isozaki et al., CZ 119:6, pp. 9991054, Fig. 1). Ga = billion years; Ma = million years; =
means equals or is the same as; AC = Accretionary Complex.

181

Gina L. BARNES

Isozaki et al. make a clear distinction between orogenic boundaries, such as the above
Tectonic Lines, and collision sutures. Two arc-arc collision sutures are named as minor
constituents of archipelagic formation: the Kzu-Matsuda fault in central
Honshu, and the West Hidaka fault in central Hokkaido,57 while only two instances of
continent-continent collision affecting Japan are identified: the major collision of the North
and South China cratons at 230Ma, and the current minor colliding of SW and NE Japan
to form the Japan Alps.58
The one large-scale over thrust sheet that resembles an Alpine nappe occurs along
the palaeo-MTL, a thrust fault that was activated between 2015Ma during Japan Sea
opening.59 Isozaki et al. state:
By utilizing mid-crustal detachment, the upper crust of the arc (the Cretaceous
batholith belt and associated pre-Cretaceous AC + meta-AC units) was horizontally
transported ocean wards. Consequently, the enigmatic occurrence of granite batholith
unit over the coeval high-P/T meta-AC in western Shikoku was achieved.60
By splitting the continental unit (the Japan arc) and over-thrusting one part onto another,
the granite batholith unit detached from its basement became rootless when over thrust,
thus conforming to the formal definition of a nappe: thrust sheets which have moved more
than about 10 km relative to the footwall.61
Another example of the differing interpretations by Charvet and Isozaki et al. concerns
the status of the North Shimanto Belt.62 Isozaki et al. identify it as an AC formed 80Ma
which partially underwent high pressure metamorphism between 7060Ma; 63 as the
lower crust of a forearc zone, it detached and slid under the main islands along the BTL,
a previous subduction plane. Charvet refers to the North Shimanto Belt as a
continent-type block, which collided and under thrust along the BTL between 80 and
60Ma. However, he does not state where this Shimanto block came from and does not
define what continent-like means in terms of the North Shimanto Belt which is an AC.
Part of the problem here is whether the docking of the North Shimanto belt is
viewed as accretion or collision. Is this perhaps a false dichotomy? Cloos suggests that
the terms collision and collider should be reserved for bodies whose subduction caused
or would cause some kind of rearrangement in the pattern of plate motion, generally the
initiation of a new subduction zone and the creation of mountains.64 These changes only
happen in continent-continent collision, which does not occur in peripheral orogens, such
as the Japanese landmass, permanently facing the open ocean at the edge and not being approached by another continental mass.65 Island arcs and other oceanic bodies do, in a sense,
collide with a continental edge (like the Izu Arc is doing now), but since they do not cause
57
58
59
60
61
62
63
64
65

182

Isozaki et al. 2010, p. 95.

Isozaki et al. 2010, pp. 83, 93.
Yanai, Aoki and Akahori, CZ 119:6, pp. 10791124.
Isozaki et al. 2010, Figure 9 caption.
Ramsay and Huber 1987, p. 521.
Charvet 2013; Isozaki et al., CZ 119:6, pp. 9991053.
Isozaki et al. 2010, p. 86.
Cloos 1993, p. 734.
Nance and Murphy 1994, p. 51.

Origins of Japanthe Big Picture Revisited

the dramatic changes listed here, it is better to consider them cases of accretion, forming
accretionary orogens.
In sum, the very mechanisms Charvet proposes for causing nappe formation (arc-arc
and continent-continent collision) are already recognized as instances of collision, and those
that are not so recognized (e.g. the Tanzawa and Misaka blocks) can be accommodated under the umbrella of accretionary tectonicswhere things do collide but are accreted without
rearrangement of plate motion or initiation of a new subduction zone or formation of ACs.
If other small blocks have arrived to cause AC formation, then where are these blocks
now? It seems the issue of terminology surrounding collision must be revisited vis--vis the
mathematical calculation of stress and strain in fault systems within structural geology, and
a clearer explanation of how accretionary prisms become Accretionary Complexes.
Conclusions
The worldwide presence of Accretionary Complexes has been recognized as both uncommon
in formation and rare in survival. Tectonic erosion by the subducting oceanic plate is the
more common process: the subducting plate grinds away at the continental edge, and can
push the eroded fragments into the serpentinizing mantle edge while carrying the rest of the
slab and trench debris down into the mantle. The eroded and subducted granite fragments
are newly hypothesized to consolidate into second continents, whose natural radioactive
heat causes magma plumes to form and break apart the continental crust above them.
In updating the overviews on the formation of the Japanese islands, several previous research results discarded by Japanese scholars were presented above in Table 1 together with
my assessments of new understandings and trends. Further important revisions include:
five Japan arc-granite batholiths have been detected; the Hida and Oki Belts are now both
acknowledged to be derived from the North China craton; the beginning of subduction in
the proto-Japan area has been pushed back to 520Ma; portions of the Qinling-Dabie-Sulu
suture have been identified in four locations in Japan; the Sanbagawa Metamorphic Belt
has been divided into two belts; the Median Tectonic Line acted in two phases, paleo-MTL
and neo-MTL which are vastly different in scale; the idea that the Outer Zone moved into
place 1000km from the south via the neo-MTL has been discarded in favour of in situ
development; and the French door model of Japan Sea opening is discarded in favour of a
drawer model.
Many of the geological belts of Japan have been conceptually reorganized as more
information about their contents and relationships is discovered (e.g. the Sanbagawa Belt
mentioned above). Table 2 provides an incomplete, but hopefully useful indication of the
natures of these belts, which in the main are correlated with Figure 1.
Acknowledgements
My greatest debt is to Professor Isozaki Yukio, not only as a prominent leader of the Miyashiro-Pacific Orogeny school of research providing all this new material, but as one who
has deigned to discuss these issues with me personally over the past decade. I thank him
also for allowing use of the geological belt map, which I have modified from the original in
his Chigaku zasshi article to remove coring borehole information and his figure references.

183

Gina L. BARNES

I am grateful to Bob Holdsworth (Durham University) for discussing thrust faults and
guiding me to further relevant readings. John Breen, as Japan Review editor, has always
supported my work and encouraged me, as an archaeologist (albeit one now holding a BSc
in Geosciences/Geology), to continue to wrestle with the difficult subject of Japanese geology. Comments, criticisms, and further reading suggestions by two anonymous reviewers
have greatly helped improve this text, but mistakes and possibly misleading colloquialisms
(boomerang effect, French door, and drawer models) are entirely my own.

REFERENCES
Barnes 2003
Gina L. Barnes. Origins of the Japanese Islands: The New Big Picture. Japan Review
15 (2003), pp. 350.
Barnes 2008
Gina L. Barnes. The Making of the Japan Sea and the Japanese Mountains: Understanding Japans Volcanism in Structural Context. Japan Review 20 (2008), pp. 352.
Charvet 2013
Jacques Charvet. Late Paleozoic-Mesozoic tectonic evolution of SW Japan: a review
reappraisal of the accretionary orogeny and revalidation of the collisional model.
Journal of Asian Earth Sciences 72 (2013), pp. 88101.
Cloos 1993
Mark Cloos. Lithospheric buoyancy and collisional orogenesis: subduction of oceanic
plateaus, continental margins, island arcs, spreading ridges, and seamounts. Geological
Society of America Bulletin 105 (1993), pp. 71537.
Isozaki et al. 2010
Yukio Isozaki, Kazumasa Aoki, Takaaki Nakama, and Shuichi Yanai. New insight
into a subduction-related orogen: a reappraisal of the geotectonic framework and
evolution of the Japanese Islands. Gondwana Research 18 (2010), pp. 82105.
Kasahara et al. 2010
Junzo Kasahara, Osamu Sano, Nobuo Geshi, Yukio Isozaki, Shigenori Maruyama, and
Shuichi Yanai. Overview of a Special Issue on Geotectonic Evolution of the Japanese
Islands under New Paradigms of the Next Generation (Part IIII). Chigaku zasshi
119:6 (2010), pp. 94758 [in English].
Nance and Murphy 1994
R. Damian Nance and J. Brendan Murphy. Orogenic style and the configuration of
supercontinents. Canadian Society of Petroleum Geologists, Memoir 17 (1994), pp. 4965.
Ramsay and Huber 1987
John G. Ramsay and Martin I. Huber. The Techniques of Modern Structural Geology,
vol. 2: Folds and Fractures. Academic Press, 1987.
Yamamoto et al. 2009
Shinji Yamamoto, Junichi Nakajima, Akira Hasegawa, and Shigenori Maruyama.
Izu-Bonin arc subduction under the Honshu island, Japan: evidence from geological
and seismological aspect. Gondwana Research 16 (2009), pp. 57280.

184